diff mbox

[v2,2/5] async_tx: Handle DMA devices having support for fewer PQ coefficients

Message ID 1486455406-11202-3-git-send-email-anup.patel@broadcom.com (mailing list archive)
State New, archived
Headers show

Commit Message

Anup Patel Feb. 7, 2017, 8:16 a.m. UTC
The DMAENGINE framework assumes that if PQ offload is supported by a
DMA device then all 256 PQ coefficients are supported. This assumption
does not hold anymore because we now have BCM-SBA-RAID offload engine
which supports PQ offload with limited number of PQ coefficients.

This patch extends async_tx APIs to handle DMA devices with support
for fewer PQ coefficients.

Signed-off-by: Anup Patel <anup.patel@broadcom.com>
Reviewed-by: Scott Branden <scott.branden@broadcom.com>
---
 crypto/async_tx/async_pq.c          |  3 +++
 crypto/async_tx/async_raid6_recov.c | 12 ++++++++++--
 include/linux/dmaengine.h           | 19 +++++++++++++++++++
 include/linux/raid/pq.h             |  3 +++
 4 files changed, 35 insertions(+), 2 deletions(-)

Comments

Anup Patel Feb. 7, 2017, 9:02 a.m. UTC | #1
On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>> The DMAENGINE framework assumes that if PQ offload is supported by a
>> DMA device then all 256 PQ coefficients are supported. This assumption
>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>> which supports PQ offload with limited number of PQ coefficients.
>>
>> This patch extends async_tx APIs to handle DMA devices with support
>> for fewer PQ coefficients.
>>
>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>
> I don't like this approach. Define an interface for md to query the
> offload engine once at the beginning of time. We should not be adding
> any new extensions to async_tx.

Even if we do capability checks in Linux MD, we still need a way
for DMAENGINE drivers to advertise number of PQ coefficients
handled by the HW.

I agree capability checks should be done once in Linux MD but I don't
see why this has to be part of BCM-SBA-RAID driver patches. We need
separate patchsets to address limitations of async_tx framework.

Regards,
Anup
Vinod Koul Feb. 7, 2017, 4:42 p.m. UTC | #2
On Tue, Feb 07, 2017 at 02:32:15PM +0530, Anup Patel wrote:
> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> > On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
> >> The DMAENGINE framework assumes that if PQ offload is supported by a
> >> DMA device then all 256 PQ coefficients are supported. This assumption
> >> does not hold anymore because we now have BCM-SBA-RAID offload engine
> >> which supports PQ offload with limited number of PQ coefficients.
> >>
> >> This patch extends async_tx APIs to handle DMA devices with support
> >> for fewer PQ coefficients.
> >>
> >> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
> >> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
> >
> > I don't like this approach. Define an interface for md to query the
> > offload engine once at the beginning of time. We should not be adding
> > any new extensions to async_tx.
> 
> Even if we do capability checks in Linux MD, we still need a way
> for DMAENGINE drivers to advertise number of PQ coefficients
> handled by the HW.

If the question is only for advertising caps, then why not do as done
for dma_get_slave_caps(). you can add dma_get_pq_caps() so that clients (md)
in this case would know the HW capability.

> I agree capability checks should be done once in Linux MD but I don't
> see why this has to be part of BCM-SBA-RAID driver patches. We need
> separate patchsets to address limitations of async_tx framework.
> 
> Regards,
> Anup
Dan Williams Feb. 7, 2017, 6:16 p.m. UTC | #3
On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>> which supports PQ offload with limited number of PQ coefficients.
>>>
>>> This patch extends async_tx APIs to handle DMA devices with support
>>> for fewer PQ coefficients.
>>>
>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>
>> I don't like this approach. Define an interface for md to query the
>> offload engine once at the beginning of time. We should not be adding
>> any new extensions to async_tx.
>
> Even if we do capability checks in Linux MD, we still need a way
> for DMAENGINE drivers to advertise number of PQ coefficients
> handled by the HW.
>
> I agree capability checks should be done once in Linux MD but I don't
> see why this has to be part of BCM-SBA-RAID driver patches. We need
> separate patchsets to address limitations of async_tx framework.

Right, separate enabling before we pile on new hardware support to a
known broken framework.
Anup Patel Feb. 8, 2017, 6:37 a.m. UTC | #4
On Tue, Feb 7, 2017 at 10:12 PM, Vinod Koul <vinod.koul@intel.com> wrote:
> On Tue, Feb 07, 2017 at 02:32:15PM +0530, Anup Patel wrote:
>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>> > On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>> >> The DMAENGINE framework assumes that if PQ offload is supported by a
>> >> DMA device then all 256 PQ coefficients are supported. This assumption
>> >> does not hold anymore because we now have BCM-SBA-RAID offload engine
>> >> which supports PQ offload with limited number of PQ coefficients.
>> >>
>> >> This patch extends async_tx APIs to handle DMA devices with support
>> >> for fewer PQ coefficients.
>> >>
>> >> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>> >> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>> >
>> > I don't like this approach. Define an interface for md to query the
>> > offload engine once at the beginning of time. We should not be adding
>> > any new extensions to async_tx.
>>
>> Even if we do capability checks in Linux MD, we still need a way
>> for DMAENGINE drivers to advertise number of PQ coefficients
>> handled by the HW.
>
> If the question is only for advertising caps, then why not do as done
> for dma_get_slave_caps(). you can add dma_get_pq_caps() so that clients (md)
> in this case would know the HW capability.

We have large number of possible capabilities for
DMA slave such as src_addr_widths, dst_addr_widths,
directions, max_burst, residue_granularity, and
descriptor_resue.

The possible capabilities of PQ offload are:
1. Number of PQ sources handled by PQ offload
(Represented by "max_pq" member of "struct dma_device")
2. Number of PQ coefficients handled by PQ offload

The above two PQ capabilities are good enough for
current PQ HW and future PQ HW so we just need a
way to specify number of PQ coefficients.

Till now all of the PQ HW always supported all 256
PQ coefficients so we never felt the need of capability
to specify PQ coefficients. The BCM-SBA-RAID is the
only HW (as far as I know) which does not support all
256 PQ coefficients.

Currently, DMAENGINE drivers use dma_set_maxpq() to
specify number of PQ sources handled by PQ HW and
Linux Async Tx uses dma_maxpq() to get number of
PQ sources.

On similar lines, we added dma_set_maxpqcoef() to
specify number of PQ coefficients and Linux Async Tx
uses dma_maxpqcoef() to get number of PQ coefficients.
If DMAENGINE driver does not specify PQ coefficients
then dma_maxpqcoef() will return 256 assuming all
PQ coefficients are supported. This approach is
backward compatible to existing DMAENGINE APIs
and will not break existing DMAENGINE drivers.

If we add dma_get_pq_caps() similar to the
dma_get_slave_caps() for PQ capabilities then we
will have to use this new method for both of the above
PQ capabilities and we have to change all DMAENGINE
drivers to use new method of specifying PQ capabilities.
I think this is too intrusive and bit overkill because its
very very unlikely to see anymore additions to
PQ capabilities.

Regards,
Anup
Anup Patel Feb. 8, 2017, 8:57 a.m. UTC | #5
On Tue, Feb 7, 2017 at 11:46 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>>> which supports PQ offload with limited number of PQ coefficients.
>>>>
>>>> This patch extends async_tx APIs to handle DMA devices with support
>>>> for fewer PQ coefficients.
>>>>
>>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>>
>>> I don't like this approach. Define an interface for md to query the
>>> offload engine once at the beginning of time. We should not be adding
>>> any new extensions to async_tx.
>>
>> Even if we do capability checks in Linux MD, we still need a way
>> for DMAENGINE drivers to advertise number of PQ coefficients
>> handled by the HW.
>>
>> I agree capability checks should be done once in Linux MD but I don't
>> see why this has to be part of BCM-SBA-RAID driver patches. We need
>> separate patchsets to address limitations of async_tx framework.
>
> Right, separate enabling before we pile on new hardware support to a
> known broken framework.

Linux Async Tx not broken framework. The issue is:
1. Its not complete enough
2. Its not optimized for very high through-put offload engines

Regards,
Anup
Dan Williams Feb. 8, 2017, 4:24 p.m. UTC | #6
On Wed, Feb 8, 2017 at 12:57 AM, Anup Patel <anup.patel@broadcom.com> wrote:
> On Tue, Feb 7, 2017 at 11:46 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>> On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>>>> which supports PQ offload with limited number of PQ coefficients.
>>>>>
>>>>> This patch extends async_tx APIs to handle DMA devices with support
>>>>> for fewer PQ coefficients.
>>>>>
>>>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>>>
>>>> I don't like this approach. Define an interface for md to query the
>>>> offload engine once at the beginning of time. We should not be adding
>>>> any new extensions to async_tx.
>>>
>>> Even if we do capability checks in Linux MD, we still need a way
>>> for DMAENGINE drivers to advertise number of PQ coefficients
>>> handled by the HW.
>>>
>>> I agree capability checks should be done once in Linux MD but I don't
>>> see why this has to be part of BCM-SBA-RAID driver patches. We need
>>> separate patchsets to address limitations of async_tx framework.
>>
>> Right, separate enabling before we pile on new hardware support to a
>> known broken framework.
>
> Linux Async Tx not broken framework. The issue is:
> 1. Its not complete enough
> 2. Its not optimized for very high through-put offload engines

I'm not understanding your point. I'm nak'ing this change to add yet
more per-transaction capability checking to async_tx. I don't like the
DMA_HAS_FEWER_PQ_COEF flag, especially since it is equal to
DMA_HAS_PQ_CONTINUE. I'm not asking for all of async_tx's problems to
be fixed before this new hardware support, I'm simply saying we should
start the process of moving offload-engine capability checking to the
raid code.
Anup Patel Feb. 9, 2017, 9:29 a.m. UTC | #7
On Wed, Feb 8, 2017 at 9:54 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Wed, Feb 8, 2017 at 12:57 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>> On Tue, Feb 7, 2017 at 11:46 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>> On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>>>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>>>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>>>>> which supports PQ offload with limited number of PQ coefficients.
>>>>>>
>>>>>> This patch extends async_tx APIs to handle DMA devices with support
>>>>>> for fewer PQ coefficients.
>>>>>>
>>>>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>>>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>>>>
>>>>> I don't like this approach. Define an interface for md to query the
>>>>> offload engine once at the beginning of time. We should not be adding
>>>>> any new extensions to async_tx.
>>>>
>>>> Even if we do capability checks in Linux MD, we still need a way
>>>> for DMAENGINE drivers to advertise number of PQ coefficients
>>>> handled by the HW.
>>>>
>>>> I agree capability checks should be done once in Linux MD but I don't
>>>> see why this has to be part of BCM-SBA-RAID driver patches. We need
>>>> separate patchsets to address limitations of async_tx framework.
>>>
>>> Right, separate enabling before we pile on new hardware support to a
>>> known broken framework.
>>
>> Linux Async Tx not broken framework. The issue is:
>> 1. Its not complete enough
>> 2. Its not optimized for very high through-put offload engines
>
> I'm not understanding your point. I'm nak'ing this change to add yet
> more per-transaction capability checking to async_tx. I don't like the
> DMA_HAS_FEWER_PQ_COEF flag, especially since it is equal to
> DMA_HAS_PQ_CONTINUE. I'm not asking for all of async_tx's problems to
> be fixed before this new hardware support, I'm simply saying we should
> start the process of moving offload-engine capability checking to the
> raid code.

The DMA_HAS_FEWER_PQ_COEF is not equal to
DMA_HAS_PQ_CONTINUE.

I will try to drop this patch and take care of unsupported PQ
coefficients in BCM-SBA-RAID driver itself even if this means
doing some computations in BCM-SBA-RAID driver itself.

Regards,
Anup
Dan Williams Feb. 9, 2017, 4:44 p.m. UTC | #8
On Thu, Feb 9, 2017 at 1:29 AM, Anup Patel <anup.patel@broadcom.com> wrote:
> On Wed, Feb 8, 2017 at 9:54 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>> On Wed, Feb 8, 2017 at 12:57 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>> On Tue, Feb 7, 2017 at 11:46 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>> On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>>>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>>>>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>>>>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>>>>>> which supports PQ offload with limited number of PQ coefficients.
>>>>>>>
>>>>>>> This patch extends async_tx APIs to handle DMA devices with support
>>>>>>> for fewer PQ coefficients.
>>>>>>>
>>>>>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>>>>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>>>>>
>>>>>> I don't like this approach. Define an interface for md to query the
>>>>>> offload engine once at the beginning of time. We should not be adding
>>>>>> any new extensions to async_tx.
>>>>>
>>>>> Even if we do capability checks in Linux MD, we still need a way
>>>>> for DMAENGINE drivers to advertise number of PQ coefficients
>>>>> handled by the HW.
>>>>>
>>>>> I agree capability checks should be done once in Linux MD but I don't
>>>>> see why this has to be part of BCM-SBA-RAID driver patches. We need
>>>>> separate patchsets to address limitations of async_tx framework.
>>>>
>>>> Right, separate enabling before we pile on new hardware support to a
>>>> known broken framework.
>>>
>>> Linux Async Tx not broken framework. The issue is:
>>> 1. Its not complete enough
>>> 2. Its not optimized for very high through-put offload engines
>>
>> I'm not understanding your point. I'm nak'ing this change to add yet
>> more per-transaction capability checking to async_tx. I don't like the
>> DMA_HAS_FEWER_PQ_COEF flag, especially since it is equal to
>> DMA_HAS_PQ_CONTINUE. I'm not asking for all of async_tx's problems to
>> be fixed before this new hardware support, I'm simply saying we should
>> start the process of moving offload-engine capability checking to the
>> raid code.
>
> The DMA_HAS_FEWER_PQ_COEF is not equal to
> DMA_HAS_PQ_CONTINUE.

#define DMA_HAS_PQ_CONTINUE (1 << 15
#define DMA_HAS_FEWER_PQ_COEF (1 << 15)

> I will try to drop this patch and take care of unsupported PQ
> coefficients in BCM-SBA-RAID driver itself even if this means
> doing some computations in BCM-SBA-RAID driver itself.

That should be nak'd as well, please do capability detection in a
routine that is common to all raid engines.
Anup Patel Feb. 10, 2017, 3:24 a.m. UTC | #9
On Thu, Feb 9, 2017 at 10:14 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Thu, Feb 9, 2017 at 1:29 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>> On Wed, Feb 8, 2017 at 9:54 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>> On Wed, Feb 8, 2017 at 12:57 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>> On Tue, Feb 7, 2017 at 11:46 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>>> On Tue, Feb 7, 2017 at 1:02 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>>> On Tue, Feb 7, 2017 at 1:57 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>>>>>>> On Tue, Feb 7, 2017 at 12:16 AM, Anup Patel <anup.patel@broadcom.com> wrote:
>>>>>>>> The DMAENGINE framework assumes that if PQ offload is supported by a
>>>>>>>> DMA device then all 256 PQ coefficients are supported. This assumption
>>>>>>>> does not hold anymore because we now have BCM-SBA-RAID offload engine
>>>>>>>> which supports PQ offload with limited number of PQ coefficients.
>>>>>>>>
>>>>>>>> This patch extends async_tx APIs to handle DMA devices with support
>>>>>>>> for fewer PQ coefficients.
>>>>>>>>
>>>>>>>> Signed-off-by: Anup Patel <anup.patel@broadcom.com>
>>>>>>>> Reviewed-by: Scott Branden <scott.branden@broadcom.com>
>>>>>>>
>>>>>>> I don't like this approach. Define an interface for md to query the
>>>>>>> offload engine once at the beginning of time. We should not be adding
>>>>>>> any new extensions to async_tx.
>>>>>>
>>>>>> Even if we do capability checks in Linux MD, we still need a way
>>>>>> for DMAENGINE drivers to advertise number of PQ coefficients
>>>>>> handled by the HW.
>>>>>>
>>>>>> I agree capability checks should be done once in Linux MD but I don't
>>>>>> see why this has to be part of BCM-SBA-RAID driver patches. We need
>>>>>> separate patchsets to address limitations of async_tx framework.
>>>>>
>>>>> Right, separate enabling before we pile on new hardware support to a
>>>>> known broken framework.
>>>>
>>>> Linux Async Tx not broken framework. The issue is:
>>>> 1. Its not complete enough
>>>> 2. Its not optimized for very high through-put offload engines
>>>
>>> I'm not understanding your point. I'm nak'ing this change to add yet
>>> more per-transaction capability checking to async_tx. I don't like the
>>> DMA_HAS_FEWER_PQ_COEF flag, especially since it is equal to
>>> DMA_HAS_PQ_CONTINUE. I'm not asking for all of async_tx's problems to
>>> be fixed before this new hardware support, I'm simply saying we should
>>> start the process of moving offload-engine capability checking to the
>>> raid code.
>>
>> The DMA_HAS_FEWER_PQ_COEF is not equal to
>> DMA_HAS_PQ_CONTINUE.
>
> #define DMA_HAS_PQ_CONTINUE (1 << 15
> #define DMA_HAS_FEWER_PQ_COEF (1 << 15)

You are only looking at the values of these flags.

The semantics of both these flags are different and both
flags are set in different members of "struct dma_device"
The DMA_HAS_PQ_CONTINUE is set in "max_pq" whereas
DMA_HAS_FEWER_PQ_COEF is set in "max_pqcoef".

When DMA_HAS_PQ_CONTINUE is set in "max_pq", it
means that PQ HW is capable of taking P & Q computed
previous txn as input. If DMA_HAS_PQ_CONTINUE is
not supported the async_pq() will pass P & Q computed
by previous txn as sources with coef as g^0.

When DMA_HAS_FEWER_PQ_COEF is set in "max_pqcoef",
it means the PQ HW is not capable of handling all 256 coefs.

>
>> I will try to drop this patch and take care of unsupported PQ
>> coefficients in BCM-SBA-RAID driver itself even if this means
>> doing some computations in BCM-SBA-RAID driver itself.
>
> That should be nak'd as well, please do capability detection in a
> routine that is common to all raid engines.

Thanks for NAKing this patch.

This motivated me to find clean work-around for handling
unsupported PQ coefs in BCM-SBA-RAID driver.

Let's assume max number of PQ coefs supported by PQ HW
is m coefs. Now for any coef n > m, we can use RAID6 math
to get g^n = (g^m)*(g^m)*....*(g^k) where k <= m.

Using the above fact, we can create chained txn for each
source of PQ request in BCM-SBA-RAID driver where each
txn will only compute PQ from one source using above
described RAID6 math. Also, each txn in chained txn will
depend on output of previous txn because we are computing
PQ from one source at a time.

I will drop this patch and send updated BCM-SBA-RAID
driver with above described work-around for handling
unsupported PQ coefs.

Regards,
Anup
diff mbox

Patch

diff --git a/crypto/async_tx/async_pq.c b/crypto/async_tx/async_pq.c
index f83de99..16c6526 100644
--- a/crypto/async_tx/async_pq.c
+++ b/crypto/async_tx/async_pq.c
@@ -187,6 +187,9 @@  async_gen_syndrome(struct page **blocks, unsigned int offset, int disks,
 
 	BUG_ON(disks > 255 || !(P(blocks, disks) || Q(blocks, disks)));
 
+	if (device && dma_maxpqcoef(device) < src_cnt)
+		device = NULL;
+
 	if (device)
 		unmap = dmaengine_get_unmap_data(device->dev, disks, GFP_NOWAIT);
 
diff --git a/crypto/async_tx/async_raid6_recov.c b/crypto/async_tx/async_raid6_recov.c
index 8fab627..2916f95 100644
--- a/crypto/async_tx/async_raid6_recov.c
+++ b/crypto/async_tx/async_raid6_recov.c
@@ -352,6 +352,7 @@  async_raid6_2data_recov(int disks, size_t bytes, int faila, int failb,
 {
 	void *scribble = submit->scribble;
 	int non_zero_srcs, i;
+	struct dma_chan *chan = async_dma_find_channel(DMA_PQ);
 
 	BUG_ON(faila == failb);
 	if (failb < faila)
@@ -359,12 +360,15 @@  async_raid6_2data_recov(int disks, size_t bytes, int faila, int failb,
 
 	pr_debug("%s: disks: %d len: %zu\n", __func__, disks, bytes);
 
+	if (chan && dma_maxpqcoef(chan->device) < RAID6_PQ_MAX_COEF)
+		chan = NULL;
+
 	/* if a dma resource is not available or a scribble buffer is not
 	 * available punt to the synchronous path.  In the 'dma not
 	 * available' case be sure to use the scribble buffer to
 	 * preserve the content of 'blocks' as the caller intended.
 	 */
-	if (!async_dma_find_channel(DMA_PQ) || !scribble) {
+	if (!chan || !scribble) {
 		void **ptrs = scribble ? scribble : (void **) blocks;
 
 		async_tx_quiesce(&submit->depend_tx);
@@ -432,15 +436,19 @@  async_raid6_datap_recov(int disks, size_t bytes, int faila,
 	void *scribble = submit->scribble;
 	int good_srcs, good, i;
 	struct page *srcs[2];
+	struct dma_chan *chan = async_dma_find_channel(DMA_PQ);
 
 	pr_debug("%s: disks: %d len: %zu\n", __func__, disks, bytes);
 
+	if (chan && dma_maxpqcoef(chan->device) < RAID6_PQ_MAX_COEF)
+		chan = NULL;
+
 	/* if a dma resource is not available or a scribble buffer is not
 	 * available punt to the synchronous path.  In the 'dma not
 	 * available' case be sure to use the scribble buffer to
 	 * preserve the content of 'blocks' as the caller intended.
 	 */
-	if (!async_dma_find_channel(DMA_PQ) || !scribble) {
+	if (!chan || !scribble) {
 		void **ptrs = scribble ? scribble : (void **) blocks;
 
 		async_tx_quiesce(&submit->depend_tx);
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index feee6ec..d938a8b 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -24,6 +24,7 @@ 
 #include <linux/scatterlist.h>
 #include <linux/bitmap.h>
 #include <linux/types.h>
+#include <linux/raid/pq.h>
 #include <asm/page.h>
 
 /**
@@ -668,6 +669,7 @@  struct dma_filter {
  * @cap_mask: one or more dma_capability flags
  * @max_xor: maximum number of xor sources, 0 if no capability
  * @max_pq: maximum number of PQ sources and PQ-continue capability
+ * @max_pqcoef: maximum number of PQ coefficients, 0 if all supported
  * @copy_align: alignment shift for memcpy operations
  * @xor_align: alignment shift for xor operations
  * @pq_align: alignment shift for pq operations
@@ -727,11 +729,13 @@  struct dma_device {
 	dma_cap_mask_t  cap_mask;
 	unsigned short max_xor;
 	unsigned short max_pq;
+	unsigned short max_pqcoef;
 	enum dmaengine_alignment copy_align;
 	enum dmaengine_alignment xor_align;
 	enum dmaengine_alignment pq_align;
 	enum dmaengine_alignment fill_align;
 	#define DMA_HAS_PQ_CONTINUE (1 << 15)
+	#define DMA_HAS_FEWER_PQ_COEF (1 << 15)
 
 	int dev_id;
 	struct device *dev;
@@ -1122,6 +1126,21 @@  static inline int dma_maxpq(struct dma_device *dma, enum dma_ctrl_flags flags)
 	BUG();
 }
 
+static inline void dma_set_maxpqcoef(struct dma_device *dma,
+				     unsigned short max_pqcoef)
+{
+	if (max_pqcoef < RAID6_PQ_MAX_COEF) {
+		dma->max_pqcoef = max_pqcoef;
+		dma->max_pqcoef |= DMA_HAS_FEWER_PQ_COEF;
+	}
+}
+
+static inline unsigned short dma_maxpqcoef(struct dma_device *dma)
+{
+	return (dma->max_pqcoef & DMA_HAS_FEWER_PQ_COEF) ?
+		(dma->max_pqcoef & ~DMA_HAS_FEWER_PQ_COEF) : RAID6_PQ_MAX_COEF;
+}
+
 static inline size_t dmaengine_get_icg(bool inc, bool sgl, size_t icg,
 				      size_t dir_icg)
 {
diff --git a/include/linux/raid/pq.h b/include/linux/raid/pq.h
index 30f9453..f3a04bb 100644
--- a/include/linux/raid/pq.h
+++ b/include/linux/raid/pq.h
@@ -15,6 +15,9 @@ 
 
 #ifdef __KERNEL__
 
+/* Max number of PQ coefficients */
+#define RAID6_PQ_MAX_COEF 256
+
 /* Set to 1 to use kernel-wide empty_zero_page */
 #define RAID6_USE_EMPTY_ZERO_PAGE 0
 #include <linux/blkdev.h>